Determinations of the absorbed dose in a 170 MeV proton beam have been performed using seven ionization chambers of different types: five cylindrical (two FWT IC-18 and three NE-2571, of which one was modified to have the central electrode made of graphite) and two plane parallel (NACP-02 and Roos FK-6). The ionization was converted into absorbed dose in the proton beam according to the generalization of the formalism provided by the IAEA Code of Practice (TRS 277), which enables the use of the same equations for all kinds of beam used in radiotherapy. The absorbed dose obtained with the two IC-18 chambers, a chamber type commonly used as a reference in proton beams, was up to 1.5% lower than that obtained with the Farmer NE-2571 chamber, which was used as the reference in this work when calibration factors in terms of NK were used. To investigate this difference, experimental ND factors for six chambers (the two IC-18 chambers, the NACP-02, the FK-6 and two of the NE-2571 chambers) were determined in a high-energy electron beam. The procedure commonly recommended for plane parallel ion chambers was used for all the chambers, using the same reference chamber, a Farmer NE-2571. In the 170 MeV proton beam all the ND factors yielded consistent absorbed dose determinations within the estimated experimental uncertainties. This finding calls into question the value of the product kattkm for the IC-18 chamber given by the IAEA Code of Practice used in this comparison, and points at possible chamber to chamber variations that theoretical kattkm factors cannot predict. The investigations enabled the determination of the Pwall(60Co) factor of the Roos FK-6 plane parallel chamber, yielding 1.003 +/- 0.5%, and a correction for the effect of the aluminium central electrode of NE-2571 chambers in proton beams, equal to 1.003 +/- 0.4%. Two of the chambers (the plane parallel FK-6 and the modified cylindrical NE-2571) were provided with calibration factors in terms of absorbed dose to water, Nw, at the quality of 60Co by the Primary Standard Dosimetry Laboratory in Germany (PTB). Using the Nw formalism excellent agreement was found with the determination based on the experimental ND, giving support to the implementation of the NW procedure in therapeutic proton beams.

Procedures for the calibration and use of plane-parallel ionization chambers in high-energy electron and photon beams have been given in the international code of practice IAEA TRS-381. In the present work, plane-parallel ionization chambers of the type PTW-34001 Roos and Scanditronix NACP02 have been calibrated using two N(K)-based procedures. For the NACP chamber the difference between the N(D,air) chamber factors determined in an electron beam and in a 60Co gamma-ray beam, respectively, is of the same magnitude as the experimental uncertainty. Results for the PTW Roos chambers, however, do not agree, in accordance with recent findings of other authors. The value determined in a 60Co gamma-ray beam is questioned and the reason for the discrepancy assigned to the correction factor for the perturbation due to the chamber wall, p(wall). New values of p(wall) have been experimentally determined by comparing absorbed dose measurements based on air-kerma and absorbed dose to water calibration procedures. A new p(wall) factor for the Roos chamber in 60Co gamma-ray beams in water (1.009+/-0.6%) was derived as the weighted average of the different determinations. The value is not significantly higher than the p(wall) factor given in TRS-381 (1.003+/-1.5%), but the combined standard uncertainty is reduced. The chamber to chamber variation for six commercial PTW Roos chambers and a Roos prototype was found to be very small.

The influence of energy and angular spread at the phantom surface on the water/air stopping-power ratios for electron beams has been studied using the Monte Carlo method. Calculated S,,,,, values for beams with most probable energies around 10 MeV and energy and angular spread of different widths as well as low-energy electron contamination have been compared with 10 MeV monoenergetic and plane-parallel beams. Variations in the calculated values of up to 1.6% at the depth of maximum absorbed dose have been obtained. The energy and angular spreads cover the interval generally occupied by clinical beams. A simple correction factor for the influence of energy and angular spread on the stopping-power ratio is proposed which only requires the dose fall-off gradient in a broad beam as input parameter. Most dosimetry protocols select water/air stoppingpower ratios as a function of the mean energy at the phantom surface and the depth of measurement. Results obtained with such a procedure have been compared with Monte Carlo calculated depth-dose curves, and corresponding slowing-down spectra and stoppingpower ratios for clinical beams. Differences in S,,,,, values smaller than 1% for beams with extreme energy and angular spread have been found. For an energy spread of 0.5 MeV and angular spread of 5", which are normal values in clinical electron beams, stopping-power ratios agree within 0.5% over the clinically useful depth interval.

The wall perturbation correction factor p(wall) in 60Co for Wellhöfer Roos-type plane-parallel ionization chambers is determined experimentally and compared with the results of a previous study using PTW-Roos chambers (Palm et al 2000 Phys. Med. Biol. 45 971-81). Five ionization chambers of the type Wellhöfer PPC-35 (or its equivalent PPC-40) are used for the analysis. Wall perturbation correction factors are obtained by assuming N(D,air) chamber factors determined by cross-calibration in a high-energy electron and in a 60Co gamma-ray beam to be equal, and by assigning any differences to the wall perturbation factor. The procedure yields a p(wall) value of 1.018 (u(c) = 0.010), which is slightly higher than the value 1.014 (u(c) = 0.010) formerly obtained for the PTW-Roos chambers using the N(D,air) method. The chamber-to-chamber variation in p(wall) for the Wellhöfer-Roos chambers is found to be very small, with a maximum difference of 0.3%. The effect of using new p(cav) values for graphite-walled Farmer-type chambers used in water in electron beams is to decrease p(wall) by approximately 0.5%. The long- and short-term stability of the Roos-type chambers manufactured by Wellhöfer is investigated by measurements at the IAEA Dosimetry Laboratory in Vienna, Austria, and at the Sahlgrenska University Hospital in Göteborg, Sweden. Calibrations made at the IAEA over several months show variations in the N(D,w) calibration factors larger than expected. based on previous experiences with PTW-Roos chambers. Measurements of the short-term stability of the Wellhöfer-Roos chambers show a marked increase in chamber response for the time the chambers are immersed in water, pointing to a possible problem in the chamber design. As a consequence of these findings, Wellhöfer is currently working on a re-design of the chamber to solve the stability problem.

The sulphuric acid used for the preparation of the Fricke dosimeter solution may contain trace impurities that can affect the yield of ferric ions. Two methods, pre-irradiation or oxidation with hydrogen peroxide, have been proposed to reduce the influence of these impurities. Fricke users sometimes omit this treatment. In the present work Fricke solutions prepared from six different brands and qualities of sulphuric acid were compared in order to study any influences of the acid on the ferric ion yield. It was shown that the use of analytical grade sulphuric acid from one manufacturer resulted in a reduction of the ferric ion yield of about 5% at an absorbed dose of approximately 20 Gy. If this solution were to be used for an absolute dose determination together with epsilon(m) G values from the literature the absorbed dose would be underestimated by the same amount.